TabNet 모델 기반 정형 데이터를 활용한 유방암 예측
Breast Cancer Prediction using Tabular Data based on TabNet Model
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 연구는 유방암 정형 데이터에 관한 딥러닝 모델 TabNet의 성능을 평가하고자 수행되었다. 연구 데이터는 WDBC(Wisconsin Diagnostic Breast Cancer)를 사용하였으며, 위스콘신 대학교에서 제공한 유방암 진단 결과 정형 데이터이다. 변수는 환자 식별 번호, 양성 여부, 종양의 상태로 이루어졌다. 어떤 변수가 유방암 예측에 가장 영향을 미치는지 파악하기 위해 정형 데이터에 특화된 딥러닝 모델 TabNet을 사용하였다. 연구 결과, 악성 종양 판단에 영향을 미치는 가장 큰 요인은 종양의 매끄러움, 질감, 반경의 평균 등이 있었다. 반면, 양성과 악성을 구분할 수 없는 요인은 비대칭적인 종양, 종양의 둘레, 오목한 점의 수 등이 있었다. WDBC에 대한 TabNet의 AUC Score에 대해서 Valid Data는 0.96356, Test Data는 0.94494로 높은 성능을 보였다.
Abstract
This study was conducted to evaluate the performance of the deep learning model TabNet on breast cancer tabular data. The research data used was WDBC (Wisconsin Diagnostic Breast Cancer), which is structured data on breast cancer diagnosis results provided by the University of Wisconsin. Variables included patient identification number, benignity, and tumor status. To determine which variables have the most influence on breast cancer prediction, we used TabNet, a deep learning model specialized for tabular data. As a result of the study, the biggest factors affecting the determination of malignant tumor were tumor smoothness, texture, and average radius. On the other hand, factors that made it impossible to distinguish benign from malignant included asymmetric tumors, tumor circumference, and the number of concave points. Regarding TabNet's AUC Score for WDBC, Valid Data showed high performance at 0.96356 and Test Data was 0.94494.
참고문헌 (14)
- Statistics on causes of death in 2021, mortality trend of malignant neoplasms (cancer) by gender, 2011-2021, "Statistics Korea", 16~ (2021)
- A. Hizukuri, "Computer-aided diagnosis scheme for distinguishing between benign and malignant masses on breast DCE-MRI images using deep convolutional neural network with Bayesian optimization", Jounal of Digital Imaging, 34, pp.116~123 (2021)
- Sercan Ö. Arik, "Tabnet : Attentive interpretable tabular learning", Proceedings of the AAAI Conference on Ariticial Intelligence, 35(8), pp.6679~6687 (2021)
- Gregory Holste, "End-toEnd Learning of Fused Image and Non-Image Features for Improved Breast Cancer Classification from MRI", Proceedings of the IEEE, 2021, 3294~3303
- Yan, "Rainfall Forecast Model Based on the TabNet Model", Water, 13(9), pp.1272~ (2021)
- Borghini, "Short Term Load Forecasting Using TabNet : A Comparative Study with Traditional State-of-the-Art Regression Models", Engineering Proceedings, 5(1) (2021)
- Shah, "Enhanced TabNet : Attentive Interpretable Tabular Learning for Hyperspectral Image Classification", Remote Sensing, 14(3), pp.716~ (2022)
- Lionel P. Joseph, "Explainable diabetes classification using hybrid Bayesian-optimized TabNet architecture", Computers in Biology and Medicine, 151(Part A) (2022)
- M. S. K. Inan, "A Hybrid Probabilistic Ensemble based Extreme Gradient Boosting Approach For Breast Cancer Diagnosis", 2021IEEE 11th Annual Computing and Communication Workshop and Conference (CCWC), pp.1029~1035 (2021)
- Haung M-W, "SWM and SVM Ensembles in Breast Cancer Prediction", PLoS ONE, 12(1) (2017)
- S. Nanglia, "An enhanced Predictive heterogeneous ensemble model for breast cancer prediction", Biomedical Signal Processing and Control, 72(Part A) (2022)
- Mohamed Hosni, "Reviewing ensemble classification methods in breast cancer", Computer Methods and Programs in Biomedicine, 177 (2019)
- Ravid Shwartz-Ziv, "Tabular data : Deep learning is not all you need", Information Fusion, 81 (2022)